To capture the resources used by short lived processes, you need to use accounting, I blogged the development of some useful code that captures data (including microstate data) from the extended accounting system here:
http://perfcap.blogspot.com/search?q=exacct Adrian On Feb 10, 2008 10:14 AM, Eugene Margulis <[EMAIL PROTECTED]> wrote: > Thank you! > > In general, am I correct in assuming that the resources of short-lived > processes that are created and die within a single sampling interval of > prstat -mL will not be reported? (i.e. the only processes/threads > reported are the ones that are alive at the end of the interval?) > -eM > > > -----Original Message----- > From: Eric Saxe [mailto:[EMAIL PROTECTED] > Sent: Friday, February 08, 2008 7:16 PM > To: Margulis, Eugene (CAR:SD10) > Cc: perf-discuss@opensolaris.org > Subject: Re: [perf-discuss] What does "/0" mean in thread name/id column > in prstat -mL output ? > > Eugene Margulis wrote: > > I often get "/0" in the PROCESS/LWPID column as part of prstat -mL > output. What does it mean? I usually get a lot of those when I just > start prstat -mL in the first screen, but I do get this after that as > well. Does this mean that the particular the process 20925 no longer > exist at the time of reporting but prstat stores its pid but not the > name? > > > > PID USERNAME USR SYS TRP TFL DFL LCK SLP LAT VCX ICX SCL SIG > PROCESS/LWPID > > 20926 root 42 9.2 3.7 0.0 2.6 0.0 0.0 42 192 14K 31K 0 cron/1 > > 20909 cemsbin 39 0.5 0.3 0.0 0.0 0.0 1.3 59 96 11K 1K 0 gzip/1 > > [b]20925 cemsbin 2.2 36 1.0 0.0 0.0 0.0 0.0 61 0 9K 87K 0 /0 > [/b] <<< > > 16371 cemsbin 36 0.6 0.3 0.0 0.0 41 3.0 19 157 6K 110 2 > java/22 > > 20921 root 26 10 0.3 0.0 0.0 0.0 4.3 59 58 7K 40K 0 ps/1 > > 16371 cemsbin 35 0.6 0.3 0.0 0.0 38 4.8 21 155 6K 228 2 > java/18 > > 16371 cemsbin 16 0.0 0.1 0.0 0.0 74 0.0 10 2K 2K 2K 0 java/6 > > > > > > This message posted from opensolaris.org > > _______________________________________________ > > perf-discuss mailing list > > perf-discuss@opensolaris.org > > > It's a process that's been zombified. You can use preap(1) to get rid of > > it if that's appropriate. > > Thanks, > -Eric > _______________________________________________ > perf-discuss mailing list > perf-discuss@opensolaris.org > _______________________________________________ perf-discuss mailing list perf-discuss@opensolaris.org
